A Study of the Design and Application of Some New Learning Rules for Neural Networks
博士 === 國立中興大學 === 電機工程學系所 === 99 === In this dissertation, we propose some learning methods including the network structure learning and network parameter learning, and then we apply the new neural network to control and synchronize a chaotic system. In the first part, a Takagi-Sugeno-Kang type self...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | en_US |
Published: |
2011
|
Online Access: | http://ndltd.ncl.edu.tw/handle/10785284256613260802 |
Summary: | 博士 === 國立中興大學 === 電機工程學系所 === 99 === In this dissertation, we propose some learning methods including the network structure learning and network parameter learning, and then we apply the new neural network to control and synchronize a chaotic system. In the first part, a Takagi-Sugeno-Kang type self-organizing fuzzy neural network is studied and an adaptive self-organizing fuzzy neural network controller system is designed based on the Lyapunov stability theory. Moreover, a proportional-integral type parameter tuning mechanism is derived. Thus not only the system stability can be achieved but also the convergence of tracking error can be speeded up. In the second part, we propose an adaptive dynamic neural network control system. The variable learning rates of the parameter adaptation laws are derived based on a discrete-type Lyapunov function to speed up the convergence rate of the tracking error.
|
---|